Freelance AI/ML Developer and 3rd-year B.Tech student building deployed AI systems — trading bots, healthcare tools, NLP pipelines. End-to-end delivery from problem definition to production. 80+ projects completed.
Background, skills, and what drives me to build
I am a 3rd-year B.Tech student in AI & Machine Learning at PCET's Pimpri Chinchwad University, working as a freelance AI/ML Developer since May 2025. I deliver production-grade solutions in Computer Vision, NLP, and Deep Learning for startup and academic clients.
With 80+ complete projects, I have developed strong practical skills handling real data, real constraints, and real deliverables. Every project runs from problem scoping to a working deployed system.
Currently focused on LLMs, MLOps, and cloud-native AI — actively seeking an AI/ML internship to contribute to large-scale production systems.
Core competencies and proficiency levels
Complete ML pipelines — training, evaluating, and deploying models for finance, healthcare, and NLP using the right algorithm for each problem.
Primary language for ML pipelines, REST APIs, data processing, and automation tools. Focus on clean, maintainable, well-documented code.
Design and train deep learning architectures for image recognition, text processing, and sequence prediction using TensorFlow and PyTorch.
Data cleaning, feature engineering, statistical analysis, and interactive dashboards — translating raw data into actionable insights.
Algorithmic trading bots with 70+ technical indicators, backtesting engines, live signal generation, and portfolio tracking dashboards.
ML models packaged as live web applications — REST APIs, Streamlit dashboards, and cloud deployment on Render and Streamlit Cloud.
80+ projects completed — 7 highlighted below
ML trading system predicting stock movement using 70+ technical indicators on a virtual $100K portfolio. Includes a full backtesting engine and live Streamlit dashboard.
Symptom-based disease checker using Jaccard similarity across 50+ diseases with ICD-10 codes. Returns top-3 diagnoses with confidence scores. Deployed on Streamlit Cloud.
Content-based recommendation engine using TF-IDF and cosine similarity across 5,000+ films. Returns similar movies in under 500ms with 82% Precision@10.
Real estate price estimator benchmarking multiple regression algorithms with feature engineering and K-fold cross-validation. Achieves R² of 0.89 on the test set.
RL agent that detects play patterns using Q-learning and Markov chains, adapting its strategy in real time after 10 rounds to maximise win rate.
Built from scratch — Three.js galaxy background, animated neural canvas, AI assistant chatbot, GitHub API integration, glassmorphism UI, and full responsiveness. Zero UI library dependencies.
Production-style REST API for personal finance management — full CRUD for transactions, user system, category filtering, pagination, and analytics (income, expenses, net balance, category breakdown, monthly summaries). Built with clean modular architecture on FastAPI + SQLAlchemy.
Concrete results from real work
Open to internships, freelance projects, and research collaborations